Nvidia Just Revealed The Future Of AI Agents In 2025..

So Jensen Kwong recently did a special address at the Nvidia AI Summit in Japan. And at this Japan summit, he actually spoke about the future of AI agents for 2025. I personally believe that 2025 is shaping up to be the year of AI agents. We've seen them from Google, we've seen them from Microsoft. Now we're seeing what Nvidia AI is cooking up with their specific AI agents and of course their entire ecosystem to help the broader community deploy and scale these AI agents for, for various different solutions.

The first AI is basically a digital AI worker. These AI workers can understand, they can plan and they can take action. Sometimes the digital AI workers are being asked to execute a marketing campaign, support a customer, come up with a manufacturing supply chain plan, optimize a chip, help us write software, maybe be a research assistant, a lab assistant in the drug discovery industry. Maybe this agent, you know, is a tutor to the CEO. Maybe there's a tutor for all of our employees. These AI, these digital AI workers, we call them AI agents, are essentially like digital employees. And just like digital employees, you have to train them. You have to create data to welcome them to your company, teach them about your company. You train them for their particular skills, depending on what function you would like them to have. You evaluate them after you're done training them to make sure that they learned what they're supposed to learn. You guardrail them to make sure that they perform the job they're asked to do and not the jobs they're not asked to do. And of course you operate them, you deploy them, provide them energy from Blackwell, the AI tokens from Blackwell, and they interact with other agents to work as a team to solve problems. Well, you're going to see all kinds of different agents. And we created several things to make it easier for the ecosystem to be able to build AI agents for companies. Nvidia is not in the service business. And we don't create, we don't deliver final product, we don't deliver solutions, but we do deliver the enabling technology to make it possible for the ecosystem to create AI, to deliver AI, to continuously improve AI. The AI agent lifecycle libraries. The Lifecycle platform is called Nemo. And Nemo has libraries for each one of the stages that I mentioned, from data curation to training, to fine tuning to synthetic data generation, to evaluation, to guard railing. And these libraries are integrated into workflows and frameworks all over the world. We're working with AI startups, service providers like Accenture and Deloitte, companies all over the World to bring this to all of the large companies. We also work with ISVs like ServiceNow so that they can create agents that use ServiceNow. Today you use ServiceNow by licensing the platform and your employees interact with the ServiceNow platform to get assistance. In the future, ServiceNow will also provide a large number of AI agents that you can rent, essentially digital employees that you can rent to help you solve problems. We're working with ServiceNow, we're working with SAP, Cadence, Ansys companies all over the world, Snowflake companies all over the world, so that we can all build agents that can be helpful to you in driving productivity in your company. Now these agents are able to understand reason, plan, take action. And these agents are collection or systems of AI models. It's not just one AI model, but a system of AI models and NEMO helps us build those. We also create pre trained AI models that we package up in what is called a NIM. And so these NIMs are microservices. They're basically AI packaged. In the old days, software was packaged in a box and they come delivered with CD ROMs. Today, AI is packaged in a microservice and inside the software is smart. You could talk to the software and you could talk to the software because it understands what you mean. And you can connect the software with other software. You can connect this AI with other AI and together you could create essentially a agent, an AI agent. So this is the first thing, let me give you an example of some of these agents.

Agentic AI is transforming every enterprise using sophisticated reasoning and iterative planning to solve complex multi step problems. AI agents help marketing campaigns go live faster with instant insights. Helping optimize supply chain operations, saving hundreds of millions in costs and reduce software security processes from days to seconds by helping analysts triage vulnerabilities. What makes AgentIQ AI so powerful is its ability to turn data into knowledge and knowledge into action. A digital agent in this example can educate individuals with insights from a set of informationally dense research papers. It was built using Nvidia AI blueprints. These are reference workflows featuring Nvidia acceleration libraries, SDKs and NIM microservices that help you rapidly build and deploy AI apps. The multimodal PDF data extraction blueprint helps build a data ingest pipeline while the Digital human blueprint provides smooth human like interactions. Hi, I'm James. A digital agent ingests PDF research papers including complex data like images, charts and tables and generates a high level summary that delivered through an interactive digital human interface. What an Exciting breakthrough in weather forecasting. The development of cordif, a new generative model, is a significant step forward in accurately predicting weather patterns. By combining a U net regression model with a diffusion model, James can also answer questions or generate new content based on the papers. Nvidia AI gives enterprises the power to automate processes, tap into real time insights and enhance workflow efficiency.

AI agents three parts Nemo, Nims and blueprints. These are all references. They're available to you in source form so that you could use it as you like and build your AI agent workforce. None of these agents can do 100% of anyone's task, anybody's job. None of the agents can do 100%. However, all of the agents will be able to do work for 50% of your work. This is the great achievement. Instead of thinking about AI as replacing the work of 50% of the people, you should think that AI will do 50% of the work for 100% of the people.

We also had Mayoshi sun actually talk about personal AI agents and I think this is going to be a really big theme for next year.

I think agents, I think every. You mentioned about enterprise AI agents differently. I support that. I'm excited about that. But also I think personal agents will really come to every one of us. Bill Gates said PC on every desktop. Steve says smartphone on your hand. I think now we should say AI agents to everybody. So each of us should have our own personal agents.

That's right.

That will help make our plans for trip vacations. Right. Education.

And it follows you your whole life.

Yeah, yeah.

Could you imagine an AI agent that knows you your whole life?

Exactly. Our grandkids, they grow up with iPhone from, you know, age one. Right. They talk and every time they see some picture, they do this even with a still picture. Right. Because they are born that, you know, with two fingers. They can, they think, you know, every picture can be blown up.

In the future they'll see a picture, they'll talk to it. Right?

They talk.

Yeah. They'll hope it's back. You can ask a question.

They will have agent, their personal agent since the age of one and their personal agent, like your, you know, second body. See you grow together with you knows everything. You are sick, you know, your health, your tutor. Yeah.

Since you were a child.

Exactly.

Yeah. Remembered everything that you read, remember everything that it taught you. Right, Exactly. Your personal Aristotle.

Totally, totally digital twin. You know, I think that is really coming now.

Remember how I said personal agents are going to be a theme for next year? If you haven't been playing around with ChatGPT's memory feature, you really do need to see exactly what information it has on you. For example, it has that my name is Andrew, I'm a male, I live in London and I eat healthily and run an AI focused channel. It does have a lot more information than that. But this allows your interactions with AI to be completely different. And this is going to be a consistent theme in 2025. Now, one of the prompts that has actually been really cool was this prompt that has been going viral. It says, Ask ChatGPT from all our interactions. What is one thing that you can tell me about myself that I may not know about myself? And I think this kind of interaction is going to completely change the game for personal AI. Now, for those of you who don't know who this is, this is Mustafa Suleiman. He's the guy currently leading the head of AI at Microsoft. And he's actually focused on building Copilot into a personal AI that can interact with you in a variety of different ways. And I think this is going to be the next AI that takes over a lot of industries.

We have the browser for distribution, we have the search engine for content, we have MSN for news. And now we have the consumer copilot, which is trying to produce this really conversational, conversational fluent AI companion, which is a search engine.

Really?

Yeah. I mean, you, you, you could think of it as a search engine, but I think it's going to evolve way beyond a search engine. I mean, this is going to become a lasting, meaningful relationship. People are going to have a real friend that gets to know you over time.

Right.

That learns from you, that is there in your corner as your support. I prefer Companion.

Companion, okay.

Which is also aspirational and not quite there yet, but it's basically getting at this idea that, you know, these AIs are really going to live life alongside you. I mean, they're going to, it's not.

Just a tool, right?

This is something that is going to end up forming a lasting, meaningful relationship with everybody over time. It is going to be your teacher, it is going to be your medical advisor, it is going to be your support network and ultimately it's going to take actions on your behalf. That's what the agentic piece is getting at.